Due to the unique characteristics of sandy soil layers, there is often a coupling effect of multiple grout diffusion patterns in the grouting process, and different slurry diffusion modes may lead to different responses of soil structures. In this study, laboratory grouting model tests were conducted with homogeneous sand under different water-to-cement (w/c) ratios to reveal the temporal variations in grouting pressure, soil stress fields, and displacement fields during the grout diffusion process. The results show that, during the grouting process in the fine sand layer, the grout mainly exhibited a compaction–splitting diffusion mode. The farther away from the grouting center, the more pronounced the hysteresis effect of soil pressure caused by grout diffusion. Meanwhile, as the w/c ratio increased, the diffusion mode between the slurry and the soil was in a transitional state. At w/c > 1.2, the primary pattern changed from the fracture–compaction pattern to the permeation–fracture–compaction pattern and fracture–permeation pattern. And the overall trend of the grouting pressure curve was similar under all of the w/c ratio conditions, showing a trend of increasing to the maximum value of the pressure first and then decreasing. With the increase in the water–cement ratio, the overall value of the grouting pressure curve showed a decreasing trend, the pressure value increased more slowly with time before reaching the maximum value, and the more obvious the influence of water–cement ratio was when w/c > 1.2. Additionally, the surface displacement also exhibited an overall decreasing trend, and it had no obvious lifting value under the condition of w/c = 1.6.

During shield tunnel construction, karst development along the tunnel axis and in the surrounding area frequently poses a significant threat to engineering safety. To address this challenge, this study proposes multiple grouting systems and systematically analyzes the key mechanical properties of five grout formulations through orthogonal experiments, identifying the optimal formulations for engineering applications. A predictive model was established using linear regression, and its accuracy was validated through independent single-factor experiments. The results indicate the following: (1) Water content is the primary factor influencing fluidity, with its significance varying by system composition. The lake mud-cement grout exhibits the highest compressive pstrength. Moderate sand addition enhances strength, but excessive amounts significantly reduce fluidity. Additives demonstrate system dependency: HY-4 effectively improves fluidity, while sodium silicate significantly increases strength but reduces fluidity. (2) The developed model demonstrates good goodness of fit, with coefficients of determination (R²) ranging from 0.74 to 0.95, without significant autocorrelation or multicollinearity. Validation experiments confirm the model’s high predictive accuracy, with overall trends consistent with the measured data. (3) The lake mud-cement grout (A3B1C3) is recommended for reinforcement projects prioritizing stability, achieving a 28-day compressive strength of 4.74 MPa. The on-site wet clay-cement grout (A2B3C1) is suitable for high-permeability formations, with a strength of 1.1 MPa and a fluidity of 292.5 mm, both exceeding standard requirements. The findings provide optimized formulations and theoretical references for grouting reinforcement in karst tunnel projects.